mm_build

Just a quick update, there is a problem with this setup. The PTFE tube is too weak to be threaded directly into the brass tube. I have had it blow out the side 4 times now. It seems to always fail at the top where the brass nozzle ends. I have been very careful threading the PTFE, but it has still failed. To make this work I would need to either reinforce the junction at the top of the tube, use a heavier tube, or most likely have something more solid and heat resistant between the brass and the PTFE. Which brings us back to PEEK.... I'm not giving up yet, but this is a step backward.

if you take the spring idea to the next step, and make an outside M7 thread all the way up to where it exits the top lasercut board (and a bit further), and put a M7 nut up there that counter acts the spring tension below... this makes the nozzle height adjustable (1 turn is 1mm)... anybody up to try it? Might be easier to thread a short plastic sleeve that fits over the tube, then glue it into place. (gluing teflon can be done apparently with a good cyanoacrylate glue (basically super glue but there are different grades))

Thanks for the tip, it works much better with an M6 thread. Preliminary testing has shown no noticeable difference from sudden changes in direction, my corners are still nice and sharp and I have cranked up Jerk and Acceleration. The spring is quite stiff, in fact maybe too stiff - when I tried to compress the spring too much the heated bowden tube stretched and broke from the pressure :(. I believe a shorter spring with less tension would be better. Also one thing I didn't test is the rapid back and forth movement you sometimes get, it shakes the printhead by default so I'm thinking it might cause the spring to resonate at certain frequencies. One other problem I encountered was some minor slippage of the spring against the heater block. The first part I printed could be seen to have a slight "wave", although it seemed to have settled into place after that. If the head was to catch on a high bit of plastic or resonate too much it might set the spring to slipping. If the spring had a proper seat that could be prevented, or else maybe some high temperature solder, or maybe even spot welded into place... I am going to try some revisions, I will update when I have something more solid. PS my bowden tube is slowly getting shorter and shorter from my modifications - time to source a new one.. It does not affect any of that..

I would suggest trying the same .stl with a different slicer, such as Skeinforge or Cura

Inspired by the ideas in this thread I have taken it a slightly different direction. I have put the M7 thread on the *inside* of the PTFE tube so that the brass tube threads into it. I have also flipped the bottom wooden plate which reverses the direction that the bowden tube is being cinched. I have placed a large spring between the heater block and the aluminum heat spreader. There is some heat transferred from the heater block to the spreader, but it is not too much since it must travel through the narrow spring, but I intend to further reduce it by adding a silicone gasket or two at the ends of the spring. Best thing about this design is that it eliminates the need for the PEEK part. This also simplifies dismantling and re-assembling the hot end such as when changing nozzles. Worst thing about this design is that the heater block does not sit flush and level by default, and the position of the spring must be manually adjusted to get it centred. Ideally there would be a recess cut for the spring to sit in at either end. I am still field testing this, no saying how it will work in the long run.

I feel it necessary to update. I have eventually concluded that my mangled nozzle tip is my own dumb fault. When I first started I was using a pair of needle-nosed vice grips to clear excess extrusion from the tip, and I am almost certain I had grabbed the tip accidentally at least once. I would just like to clear ultimakers name in case I implied that's how they send stuff out. Newbies make mistakes they don't even realize!

Thanks for the tips!

I also use a hackerspace, and dare I say the reprap ppl are slightly jealous of my ultimaker. I was up and running within a week (some teething problems), while the rep rap people took years to get to where they are... I'm printing at twice the speed and half the resolution. There are also some Makerbot people and... well let's just say they look like this -> The extruder AND the bowden/hotend connection are the weak parts of the Ultimaker design, but many people have been taking steps to making them better. It's important to realize that the ultimaker is just as open source and modifiable / hackable as any reprap design. For the amount of "hacking" you would to do get a reprap up and running, there is far less trouble to upgrade the ultimaker's extruder with whatever design you have in mind. Also, I am sure the ultimaker people are keeping an eye on what people are saying and doing, it's only a matter of time before they unveil any number of boot-strap modifications and upgrades. The community is getting larger as well, (safety in numbers). For the record I have had very little problems with my extruder or hot end. The hot end can be finicky, especially if you like to disassemble and reassemble it. But it is nothing that is insurmountable. The extruder is a pretty basic deign, but it "works" for the most part.

I'm... Obsolete! :cry:

Sounds like a problem with the PCB :( The only other thing I would double check is that there is a solid connection to the motor wires. Try removing and re-inserting all 4 motor wires, making sure the screw-clamp is fully loosened before re-inserting the bare wire into it. Also the plastic part where the wires screw into is removable, you might want to try swapping it with one next to it and testing again. Disconnect all limit switches in the mean-time just to be absolutely sure. (edit - Oh I see you already swapped the wires, well you can try swapping the plastic and see if it doesn't work when Y is plugged into it, and if you have a multimeter try getting a voltage reading off the PCB connections while the motors are in motion, compare X terminals with Y terminals)

Yes, it takes much longer for the print head to heat up, but the firmware seems to adjust to being constantly cooled and is able to output a constant temperature after it gets there. Actually I have some molded silicone which is good to 294 C, I plan on dipping the head + heater block in this, it should provide an effective thermal barrier. Will let you know how it all turns out.

Daid, have you considered a cooling setup like this? I am still making additions to the design, such as a bottom and sides. I made this because I realized that most of the air is not being pushed out the tiny hole in the fan ducts, the fans are not strong enough and you will get more airflow having the fan blow directly on the part instead of trying to bend the air around corners and force it through a narrow opening. I have soldered connectors on the two unused "12 volt" supply at the bottom of the board. The only downside so far is the fans are on constantly when the machine is powered. PS I used 50x50x10mm fans because they are lower profile than the Ultimaker supplied one which is closer to 12mm

Printing from Pronterface, I really don't want to go back to ReplicatorG... not only that but it's difficult to say since I can print for hours sometimes, and other times it cuts out after not very long I don't have a good way of diagnosing at this point.

Hello fellow Ultimakerians, I have a simple question about occasionally losing connection to my machine during a print. This is a fairly recent problem, and my intuition tells me it's something to do with windows (or an application) poking around in the USB channel when it shouldn't be. My long term solutions will include the Ulticontroller, and possibly a separate PC / windows partition that is whistle clean. Anyone else have this problem or has found a fix?

This has already been done! I don't have any links handy but I'm sure you can find it in a search or someone more familiar with it can chime in. The basic problem with it is, those flexible shafts have too much play and depending on start/stop/retraction, it is difficult to model where you actually are in terms of rotation from the one end to the other. People have used laser diodes to track this with software and make compensations, needless to say it removes some of the desired simplicity...

I can't post to the google group either (usually, it's very random... not worth my time) Pls someone correct me if I am wrong since I also would like to know more, but whether your stepper is in microstepping or not has nothing to do with the current "speed" at which you are travelling. As far as I understand there is a jumper you must set which determines which mode you are in, and and it does not change with speed at all. Most motors (ie - all) have better torque at lower speeds, and microstepping also lessens torque overall, but these are separate things.

Ian, thank you so much for doing this. Forum was in desperate need of a clean-out. If you have any other powers or contact with those that do, Pls pls, can you make it harder for the spammers to start? It really shouldn't be that hard to prevent them in the first place. (PS This forum has felt really neglected over the past months despite its increasing user base.)

This looks very interesting, thank you for sharing. Pls keep us up to date.

There's already 2 topics on the front page about this (at the time of your posting). If anyone wants to buy one, feel free! it is at your own risk though.

Actually, I had this problem for the longest time and never did figure out what was causing it. It just sort of went away after a while. I suppose it had something to do with the firmware. I still have some of my old "rectangular" test cubes. I would suggest upgrading your firmware to Marlin.

Your problem looks very similar to a problem I have. In my case it is caused by the end of the brass nozzle not being flat, but cut at an angle. I have not gotten around to fixing it yet, but I plan on sanding it , I will also need a 0.4mm plunge (such as a drill bit) in order to make sure there are no burs left inside the nozzle after sanding. This is more easily visible with the eye. Some feed/flow settings will make the problem more or less obvious.

SUCCESS-ish! Since the voltage at X2-1 was different from the voltage at pins 5/6, I thought perhaps there was a break or a short or something. Well, I directly soldered pins 5/6 to X2-1 with a bit of wire, and the result so far is desirable. My machine is reporting accurate temperatures for the first time since the trouble began. I still don't know what the phantom voltage coming from PCB side was, but connecting directly to pins 5 and 6 seems to override whatever was coming off the main PCB. Perhaps it was a signal used to search for the AD597? At any rate things are moving forward now, I will update if I encounter any more problems.

Both places, X2-1 when the signal wire is connected to the main PCB. When the signal wire is disconnected, it's only at the main PCB connector. The "bogus signal" does not appear to travel backwards to the pins on the AD597, for whatever reason.

I have not had much time to play with this lately, but I finally replaced the AD597 and did some testing. The AD597 chip itself appears to be functioning correctly, however I am still getting a bogus signal. Measuring voltage between pins 5/6 and ground, the output appears to be bang-on correct to 10mV/C. Heating the extruder in pronterface and measuring the voltage at the pins shows that the mV are increasing and decreasing proportionally to the temperature. The problem is that when I measure from the green signal wire (where pins 5 and 6 lead to), the voltage is pretty random. I am measuring anywhere between .45 and 4.5 volts at the signal wire. It seems to change when I reset the machine, and it seems to be totally random. However, this voltage reading is consistent with the temperature reading in pronterface. For example, when the voltage is about 4.5, I will get a temp reading of about 450 degrees. The voltages in the signal wire are not coming from the TC board. I completely unplugged the TC board and measured the signal pin on the main PCB, and the voltage was the same. Am I supposed to be getting random voltages from this pin at the PCB side? And if not any tips for what next? Thanks.

Verify that the small belts connected directly to the steppers are tight, if they do not already have washers they really should. Also, does it print squashed holes both vertically (through the z layers), and also horizontally (x and y laying flat)?